| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Minireview |
Women’s Health and Musculoskeletal Biology, Wyeth Research, Collegeville, Pennsylvania 19426
Address all correspondence and requests for reprints to: Heather Harris, Women’s Health and Musculoskeletal Biology, Wyeth Research, 500 Arcola Road, RN3163, Collegeville, Pennsylvania 19426. E-mail: harrish{at}wyeth.com.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONSUMMARY OF AVAILABLE TOOLS...ER{alpha} MEDIATES THE MAJORITY...ROLES FOR ERß IN...UNANSWERED QUESTIONS ABOUT...CONCLUDING REMARKSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONSUMMARY OF AVAILABLE TOOLS...ER{alpha} MEDIATES THE MAJORITY...ROLES FOR ERß IN...UNANSWERED QUESTIONS ABOUT...CONCLUDING REMARKSREFERENCES |
|---|
. This second receptor, designated ERß (4), was unexpectedly found while searching for additional androgen receptors using a rat prostate cDNA library and degenerate PCR primers. It would not be an overstatement to say that this discovery electrified the field of estrogen biology, causing a reexamination of all our understandings and assumptions about the mechanism of estrogen action.
Early studies, naturally, focused on cloning ERß from additional species (5, 6) and describing ERß’s ligand specificity (7, 8, 9) and tissue distribution (10, 11, 12, 13). Two key conclusions can be drawn from this body of work: First, 17ß-estradiol is the most potent endogenous ligand for ERß, and it binds equally well to ER and ERß. Second, ERß is widely distributed, although it is not highly expressed in the uterus. One logical extrapolation of this information is that a selective ERß ligand would be expected to mimic estradiol’s actions in a variety of target tissues but would have less impact on the uterus. This was the premise behind part of the drive to identify/design selective ligands: not only could these compounds help dissect ER from ERß function but they might also carry clinical benefits. Early speculation about potential clinical benefits was primarily related to the treatment of menopausal disorders such as osteoporosis, cardiovascular disease, cognition, and urinary incontinence (14). If this list seems expansive and unduly ambitious, recall that ERß’s discovery came at a time when estrogen therapy in postmenopausal women was thought (at least at a research level) to potentially confer benefits on almost all organ systems and was prescribed clinically for long-term use.
This article will review the current state of knowledge about the function of ERß and will be biased toward discussion of recent in vivo data generated from ER and ERß knockout (KO) mice as well as a summary of the preclinical characterization of ER- and ERß-selective agonists. A number of other reviews on ERß have recently been published that examine estrogen action in general and lessons learned from in vitro studies (15, 16, 17).
|
SUMMARY OF AVAILABLE TOOLS AND CAVEATS |
|---|
|
TOPABSTRACTINTRODUCTIONSUMMARY OF AVAILABLE TOOLS...ER{alpha} MEDIATES THE MAJORITY...ROLES FOR ERß IN...UNANSWERED QUESTIONS ABOUT...CONCLUDING REMARKSREFERENCES |
|---|
ERKO mice exist. The first mouse constructed (Ref. 3 ; designated ERKOCH in this review) is still the most widely used, but residual effects of estrogens in this animal cannot be automatically attributed to ERß because of low levels of ER splice variants that retain some activity (18). More recently, an apparently complete KO of ER has been accomplished, and these mice will be designated ERKOST (19). Three groups have independently constructed mice lacking ERß, and there is disagreement in the field as to their phenotype. In this article, ßERKO mice will be designated as follows: ßERKOCH for the mice produced in Chapel Hill by Krege et al. (20) that are commercially available from Taconic Farms, Inc. (Germantown, NY); ßERKOKI for the colony of these mice that were subsequently established at the Karolinska Institute; ßERKOST for the mice produced by Dupont et al. (19) in Strasbourg, France; and ßERKOWYE for the mice produced by Wyeth (21).
Although independently generated KO mice might not be expected to exhibit identical phenotypes, the issue confronting scientists studying ßERKO mice is unexpected in that investigators studying ßERKOCH and ßERKOKI observe markedly different phenotypes (detailed examples discussed below), whereas the ßERKOST and ßERKOWYE mice seem similar to the ßERKOCH mice. Potential explanations for the phenotype differences have been discussed but not rigorously tested. One hypothesis is that factors present in the ßERKOKI environment elicit the phenotypes seen. These might include dietary differences or the presence of endemic pathogens. This hypothesis is readily testable by cohousing ßERKOCH mice with ßERKOKI mice in the Karolinska Institute’s colony, but this has not been done as yet. Although the data generated from the ßERKOKI mice are dramatic and provide compelling evidence for a role for ERß in a variety of diseases, it is very puzzling that similar observations have not been made in other ßERKO mice. As with the selective compound data discussed below, and indeed with any scientific observation, the dataset inspiring the highest confidence is obtained from independently corroborating observations.
Selective Small Molecules
A number of selective ERß agonists have been described in the peer-reviewed scientific literature (see Ref. 22 for a comprehensive review including patent sources), but a minority has been evaluated extensively in vivo. These include DPN (23), ERB-041 (24), WAY-202196 (25), WAY-200070 (24), and 8ß-VE2 (26) (Fig. 1
). Selectivity of these compounds was measured in in vitro screens using transcriptional assays and/or competitive radioligand-binding assays. All of these compounds are at least 70-fold selective for ERß over ER (based on a radioligand-binding assay), and their potencies range from 18–180% that of 17ß-estradiol (Table 1).
|
|
agonists have been described and characterized in vivo. These are PPT (27, 28) and 16-LE2 (26, 29), and their structures are shown in Fig. 1
. In some publications, 16-LE2 is known as Cpd1471 (30). In competitive radioligand-binding assays, these compounds are at least 250-fold selective for ER and have potencies of about 50–60% that of 17ß-estradiol (Table 1). Taken together with the ERß-selective agonists, these reagents are powerful tools to elucidate the respective functions of ER and ERß. Although good comparative data exist for both types of ER-selective compounds on classic estrogenic endpoints and the results are, by-and-large, consistent (see below), relatively few studies have been published on the ERß agonists’ positive in vivo activity, and few direct comparisons have been made. Therefore, it is premature to make many statements about ERß class effects that will be shared among all members. Just as selective ER modulators can vary in their efficacy on a given target, it is to be expected that even bona fide selective ERß agonists will exhibit a range of activities. Moreover, typically these ERß-selective compounds were classified as agonists based on their profile in (an often) contrived in vitro model or using three-dimensional x-ray cocrystal structure. It is entirely possible that the compounds described to date do not elicit the full range of biological activities able to be exhibited by ERß. Thus, the accurate definition of ERß-mediated activities requires the simultaneous testing of a wide range of highly selective ligands in a variety of in vivo models.
|
ER MEDIATES THE MAJORITY OF ESTROGEN’S EFFECTS ON CLASSIC ESTROGEN TARGET TISSUES
|
|---|
|
TOPABSTRACTINTRODUCTIONSUMMARY OF AVAILABLE TOOLS...ER{alpha} MEDIATES THE MAJORITY...ROLES FOR ERß IN...UNANSWERED QUESTIONS ABOUT...CONCLUDING REMARKSREFERENCES |
|---|
-ethinyl-17ß-estradiol, have well-characterized and familiar effects on several organ systems. These include stimulation of the uterus, facilitation of proliferation/endbud development in the mammary gland, control of ovulation, maintenance of bone mineral density, and negative feedback to the hypothalamus and pituitary. As has been reviewed extensively elsewhere (16, 31), mice not expressing ER lack many of these responses. Although the published literature is not unanimous on the subject, on balance, the data suggest that ERß plays a minor role in several organs/organ systems that are highly responsive to estrogens. These are discussed in more detail below.
Uterus
ERKOCH mice have well-documented defects in uterine responses to estrogens (reviewed in Ref. 31). However, nonselective estrogens cause an equivalent increase in uterine wet weight in wild-type and ßERKOCH and ßERKOWYE mice (Korach, K. S., and H. A. Harris, unpublished observations). Moreover, microarray studies indicate that both the early and late genomic responses to estradiol are indistinguishable between wild-type and ßERKOCH mice (32). In contrast to these findings, the ßERKOKI mice exhibit an enhanced response to estrogenic stimulation as measured by luminal fluid production and expression of complement C3 protein (33).
The data gathered from selective ER agonists are consistent with the KO mouse phenotype, suggesting a minor role for ERß in uterine physiology. ER appears to be necessary and sufficient to mediate estrogen’s increase in wet weight (at least in rats) as evidenced by the fact that PPT (28) and 16-LE2 (26) are as efficacious as reference estrogens on this endpoint. In terms of ERß-selective compound activity, uterine weight has been measured for all five selective agonists. DPN, ERB-041, WAY-202196, and WAY-200070 do not significantly increase uterine weight at 
50 mg/kg when given sc to rats and/or mice (Refs. 24 and 25 ; and Harris, H. A., unpublished observations). The steroidal selective agonist, 8ß-VE2, increased uterine weight at a sc dose of 0.5 mg/kg. It is interesting to note that the in vivo selectivity of 8ß-VE2 is more in line with expectations based on its in vitro selectivity than the other compounds, which appear more selective in vivo that what would be predicted from their in vitro potency and selectivity. There are potential explanations for this observation. First, there is not an exact relationship between binding and activity. Second, these compounds are likely differentially bound to plasma proteins and/or metabolized.
Mammary Gland
Data from ERKOCH, ßERKOCH, and transgenic aromatase/ERKOCH mice clearly indicate that ERß is not required for normal mouse mammary gland development or function (16, 34). In contrast, ßERKOKI mice have enlarged alveoli and reduced expression of several markers of differentiation and a concomitant increase in the proliferation marker Ki67 (35). Little published data currently exist on the effects of selective ER agonists on estrogen-dependent functions in the mammary gland.
Hypothalamic/Pituitary Effects
In contrast to ERKOCH mice, ßERKOCH mice have normal circulating levels of LH and estradiol (36), suggesting that ER controls the negative feedback of estrogens to the brain (16). Similar to estradiol, 16-LE2 suppressed LH and FSH production in the ovariectomized rat, again showing ER sufficiency in mediating this negative feedback loop. When administered at nonuterotrophic doses, the ERß-selective agonists DPN, WAY-200070, and 8ß-VE2 had no effect on FSH and/or LH nor did they inhibit ovulation, suggesting the inability of ERß to influence this axis (26, 29, 37, 38).
Another well-known effect of estrogens on the pituitary is the stimulation of prolactin secretion. In mice, PPT increased prolactin levels whereas DPN had no effect. It should be noted that the efficacy of PPT was less than that of estradiol at the dose tested, but that a full dose-response study was not performed (39).
Skeleton
Careful characterization of ER null mice has shown a variety of phenotypes, suggesting that both ER and ERß play a role in bone development. However, these effects are not seen at all ages in both genders or all investigators’ ßERKO mice (see Ref. 36 for review). Moreover, estradiol protected against ovariectomy-induced bone loss in ßERKO mice, indicating that ERß is not necessary for this activity. Extrapolation of these findings in rodents to effects in humans is problematic; however, an unambiguous role was established for the necessity of ER in normal bone development when a human male was discovered with an inactivating mutation in this receptor (40). This individual had unfused epiphyses and markedly reduced bone mineral density and was unresponsive to estradiol therapy.
The activity of selective ERß agonists supports the idea that ERß is not a major player in estradiol’s maintenance of adult skeletal mass. As with the uterine responses, selective ER stimulation with PPT or 16-LE2 is as efficacious as a reference estrogen at preservation of bone mineral density after ovariectomy in rats (26, 28), whereas ERB-041, WAY-200070 or 8ß-VE2 were inactive at nonuterotrophic doses (24, 26, 41).
Ovarian Physiology
ERß is highly expressed in the granulosa cell, and all types of ßERKO mice are subfertile. Ovaries from both ßERKOCH and ßERKOST mice have fewer corpa lutea than their wild-type counterparts but only the ßERKOCH mice showed an increase in the number of atretic follicles (19, 20). Under gonadotropin stimulation, ßERKOCH and ßERKOST mice produced fewer oocytes, although the severity of impairment varied and likely represented partial compensation in some animals by ER. These data indicate a role for ERß in folliculogenesis and are supported by the activity of 8ß-VE2 in the hypophysectomized rat (29).
Potential for ERß-Selective Compounds to Block Estradiol’s Activity
Given the inactivity of selective ERß agonists in classic estrogen target tissues, several studies have examined the potential for these compounds to block a reference estrogen’s effects. In the uterus, when WAY-200070, ERB-041, or 8ß-VE2 are coadministered with 17ß-estradiol, these compounds do not block the estradiol-mediated increase in uterine weight (24, 26, 41). Additionally, WAY-200070 did not block the antiresorptive effects of 17ß-estradiol in rat models of osteopenia (24), and 8ß-VE2 was unable to block 17ß-estradiol’s effects on pituitary (gonadotropins) or liver (angiotensin I, IGF-I, high density lipoprotein, or total cholesterol) responses (26). DPN did not antagonize PPT effects on LH, FSH, or prolactin (37).
Frasor et al. (39) have found that DPN has the potential to mildly antagonize the uterine effects of PPT. In the mouse, coadministration of DPN with a modest dose of PPT led to a 30% reduction of PPT-induced wet weight. Additionally, mild inhibitory effects were seen on some mRNA markers of estrogen action in the uterus. On the other hand, DPN had modest stimulatory effects on the activity of PPT on other markers of estrogen action. It remains to be determined whether these effects indicate a subtle modulatory effect of ERß on uterine response to estrogens or if they are observations specific to DPN.
|
ROLES FOR ERß IN NONREPRODUCTIVE ORGAN SYSTEMS |
|---|
|
TOPABSTRACTINTRODUCTIONSUMMARY OF AVAILABLE TOOLS...ER{alpha} MEDIATES THE MAJORITY...ROLES FOR ERß IN...UNANSWERED QUESTIONS ABOUT...CONCLUDING REMARKSREFERENCES |
|---|
Prostatic Actions
The role of estrogens in prostatic physiology is controversial. Historically, high doses of estrogens were used to treat prostate cancer, but it was thought to act via down-regulation of gonadotropins. When ERß was discovered in the prostate epithelium, it stimulated investigation of direct interactions of estrogens on the prostate. Although neonatal exposure to high amounts of estrogens alters prostate development, two studies using KO mice (ßERKOKI, ßERKOCH, and ERKOCH) clearly show this imprinting is mediated by ER (42, 43).
Older ßERKOKI mice have hyperplastic prostates (44), a phenotype that is not seen in ßERKOCH (42, 45) or ßERKOST (19) mice. In addition to abnormal histological changes in these mice, markers of proliferation are increased and androgen receptor (protein) expression is increased. It appears that the prostate locally produces a dihydrotestosterone metabolite that is ligand for both ERs: 5-androstane-3ß,17ß-diol (44). Exogenous administration of this compound to mice decreased markers of cell proliferation in the developing prostate of wild-type mice and this effect was absent in ßERKOKI mice (46). However, as pointed out by Taylor et al. (47), because the prostate is so androgen sensitive, it is difficult to separate direct effects of estrogens on this organ from their ability to modulate androgen action. Accordingly, this group used organ cultures of prostates isolated from rats (newborn) or wild-type ERKOCH, and ßERKOCH mice (postnatal d 1). In cultures of both rat and mouse prostates, testosterone supports growth and branching, whereas coadministration of estradiol reduced growth and promoted apoptosis, especially in the distal epithelial cells (areas of active branching). These proapoptotic effects were seen in both ERKOCH and ßERKOCH mice, suggesting neither ER is involved. Because double KO mice were not available, selective agonists were used to try and mimic estradiol’s effects. However, neither PPT, DPN, nor 5-androstane-3ß,17ß-diol could replicate the effects of estradiol on the explants, although they each had other effects on prostate morphology. It should be noted that although low levels of testosterone were required to maintain tissue development, very high concentrations of estradiol were used to see the proapoptotic effects (15 µM). These concentrations are similar to those used for neonatal imprinting studies but may have little relevance to adult exposures. Further analysis of the estradiol-treated prostates showed that ERß protein was up-regulated, and androgen receptor protein was down-regulated in distal areas of prostates, regardless of genotype. Thus estrogens, especially at pharmacological doses in the developing organ, may impact prostate physiology independent of either ER.
Vasomotor Instability
Estrogens are frequently prescribed for the relief of hot flushes in peri- or postmenopausal women. Both ERs are expressed in the hypothalamus, the thermoregulatory center of the brain; therefore it is possible that estrogen’s suppression of vasomotor instability could be mediated by ERß and/or ER. This question has been investigated using both KO mice and selective ER agonists, and studies agree that activation of ER is sufficient to modulate tail skin temperature in the rodent, an animal model of hot flush. First, PPT was as effective as ethinyl estradiol at reducing tail skin temperature changes induced by acute morphine withdrawal in ovariectomized rats (28). Second, Opas et al. (48) found that estradiol cypionate was as effective at suppressing ovariectomy-induced tail skin temperature increase in ßERKOCH as wild-type mice. The data on whether ERß activation has a similar effect are mixed. In the Opas et al. study mentioned above, estradiol cypionate also worked well in ovariectomized ERKOCH mice, whereas WAY-200070 was inactive in the morphine model (24). Additional studies are needed to resolve this apparent inconsistency as to the role of ERß in mediating estrogen’s actions on hot flushes.
Cardiovascular System
Estrogens have well-appreciated effects on the cardiovascular system in preclinical models, but human clinical trial data have not been unanimously supportive. Reasons behind this apparent discrepancy are under investigation (see Ref. 49 for review) and may depend on the age at which treatment is given (50). Despite this controversy, enthusiasm for defining the role of the ER subtypes in mediating estrogen’s preclinical effects remains high. KO mice and selective ER agonists have been used to investigate the relative contributions of ERß and ER in models of vascular injury, vascular function, and cardiac injury.
After carotid artery denudation, vascular smooth muscle cells proliferate, leading to a narrowing of the vessel lumen. Estradiol can prevent this neointima formation in both rats and mice. Early studies using ERKOCH mice suggested estradiol might work via ER or ERß to prevent intimal thickening, but more recent data from ERKOST mice suggest ER is the sole receptor responsible (51). No highly subtype-selective agonists have been tested in this model; thus it remains to be seen whether these tools will confirm observations from the KO mice.
Several studies have examined ER subtype effects on vascular function. Estradiol does not reduce phenylephrine-induced contraction in endothelial-denuded aortic vascular rings of ßERKOCH mice as it does in wild-type mice, and this defect is due to reduced inducible nitric oxide synthase produced (52). In keeping with these results, rat aortic rings with intact endothelia precontracted with noradrenaline relaxed significantly when treated acutely with nanomolar levels of estradiol or PPT, whereas DPN had no significant effect over a concentration range expected to be selective for ERß (53). The PPT effect was blocked by ICI-182780, strongly suggesting an ER-dependent effect. These data suggest ER mediates relaxation effects but is in contrast to the work reported by others (54, 55) that use PPT and DPN to support a role for both ERs. However, these other studies used very high (>1 µM) concentrations of compounds that are clearly pharmacological and beyond the range where these compounds are ER or ERß selective in in vitro assays (23, 27). A study in the spontaneously hypertensive rat confirms an obligatory role for ER in mediating vessel relaxation (56). In this study, ovariectomized rats were treated for 4 wk, after which aortic rings were isolated for testing. Aortic tissue from rats treated with estradiol or 16-LE2 (Cpd1471) and precontracted with KCl was more responsive to acetylcholine-induced relaxation than tissue from ovariectomized rats. These effects were mediated by the endothelium, as contraction/relaxation profiles of denuded aortic rings were not different between the treatment groups. Moreover, endothelial nitric oxide synthase was up-regulated by treatment with estradiol and 16-LE2, again pointing to an ER-dependent effect on the endothelium.
Three studies have looked at in vivo effects of ER subtypes in models of cardiac injury. Korte et al. (57) looked at the effect of ER or ERß deletion on various cardiac functions after anterior myocardial infarction. Genotype did not influence mortality or infarction size [in contrast to the findings of Pelzer (58) discussed below]. Moreover, the ERKOST mice had no defects in ECG parameters. In contrast, the ßERKOCH mice had prolonged QT, QTc, JT, and JTc after infarction when compared with wild-type mice (regardless of injury) and sham-operated ßERKOCH controls. Moreover, there was a dramatic ERß-dependent reduction in ventricular premature beats. These phenomena may be attributed to decreased expression of the potassium channel Kv4.3 mRNA seen in the ßERKOCH mice.
In a similar study of chronic anterior myocardial infarction, ßERKOCH mice had lower survival rates, retained more fluid, and had elevated levels of ventricular pro-atrial naturetic peptide than their wild-type counterparts. Thus, although the data are not entirely concordant, both studies conclude that ERß plays a role in the response to myocardial infarction. Future studies will likely test whether selective ER agonists are protective in this model.
Finally, in an ex vivo ischemia/reperfusion model using hearts excised from ERKOCH and ßERKOCH mice and treated with isoproterenol, Gabel et al. (59) found that female hearts sustained less injury (as measured by a reduction in infarction size) than males, and that this was attributable to ERß. Expression profiling suggested that ßERKOCH females had reduced levels of various genes involved in fatty acid metabolism, thus perhaps providing a mechanistic explanation.
When the heart is subjected to increased pressure via aortic constriction, cardiac hypertrophy develops. Skavdahl et al. (60) observed that male mice developed larger hearts than females and further discovered, using ERKOCH and ßERKOCH mice, that ERß was responsible for this attenuation. However, a study in spontaneously hypertensive rats indicates that stimulation of ER is protective (30). These rats develop cardiac hypertrophy as a consequence of chronically elevated blood pressure. Although estradiol or 16-LE2 treatment for 12 wk did not lower blood pressure, both compounds reduced heart weight and improved cardiac function. These effects were blocked by ICI-182780, indicating the activity depends on an ER. It remains to be seen whether ERß agonists can also achieve this effect. It will also be interesting to determine the effect of selective ERß agonists on blood pressure because the ßERKOCH mice have hypertension (52).
Brain and Behavior
ERß is expressed throughout the brain (see Ref. 10 for review), and there has been much interest in defining its function. Specific areas of investigation have been stroke/neuroprotection, anxiety/depression, and learning/memory.
Estrogens protect against neuronal cell loss in a variety of ischemic stroke models. These models differ in whether the tissue is reperfused and the level of estradiol required for protection; thus the data are not entirely comparable. Early data using ßERKOWYE and ERKOCH mice indicated an obligatory role for ER (61) in preserving neurons after middle cerebral artery occlusion leading to permanent cerebral ischemia. More recently, using a 15-min global ischemic insult, followed by reperfusion, DPN significantly reduced neuronal damage in the caudate nucleus and CA1 pyramidal cell layer (62) in mice. A similar study by Miller et al. (38) found both WAY-200070 and PPT protected the CA1 region of the hippocampus although there was a pronounced bimodal distribution among the animals within a group.
One area of particularly active investigation is whether ERß might ameliorate anxiety and/or depression. Support for this hypothesis is derived from distribution studies as well as estradiol/ERß influences on the serotonergic system. Double immunolabeling studies show that more than 90% of ERß-expressing neurons also express tryptophan hydroxylase, the rate-limiting step in serotonin synthesis; although deletion of ERß in ßERKOCH mice does not affect its expression (63). A functional connection between these two proteins was made by the discovery that estradiol up-regulates tryptophan hydroxylase-1 mRNA via ERß in the dorsal raphe nucleus and shown by comparing estradiol’s activity in wild-type and ßERKOCH mice (64). Interestingly, such regulation may have a local effect in the midbrain, because estradiol does not affect tryptophan hydroxylase-2, the more widely expressed isoform. It remains to be shown whether this increase in mRNA translates to increased protein expression.
A number of behavioral studies have been published that use either KO mice or selective ER agonists to investigate the effects of ERß on anxiety and/or depression. In the forced swim test model of depression using mice, estradiol reduced immobility time comparably to a reference compound, desiprimine, and estradiol’s effect (but not that of desiprimine) was lost in ßERKOCH mice (65). ERKO mice have not been evaluated as yet. Rats’ performance in the forced swim test also implicates a role for ERß: immobility times were similarly reduced when rats were given estradiol, DPN, and desiprimine, but PPT had no effect (66).
In a model measuring anxiety, Krezel et al. (67) reported an increase in anxiety-related behaviors of ßERKOST female mice in the elevated plus maze or in the open-field test, perhaps attributable to abnormalities in the amygdala. These behavioral observations were confirmed by Inwalle et al. (68) using ßERKOCH mice, and it was further reported that estradiol did not influence the behavior of either genotype. Two other studies used rats and measured the time they spent in a lighted open field or the open arm of an elevated plus maze (69, 70). Estradiol and DPN clearly increased time spent in the open area, and PPT had no effect although it affected other behaviors. Moreover, these antianxiety effects of estradiol and DPN were effectively blocked by coadministration of tamoxifen, suggesting an ER-mediated effect.
ERß’s role in synaptic plasticity and learning is also being investigated (71). Male and female wild-type and ßERKOWYE mice were trained to associate a shock with a particular context (operant chamber) and cue (tone). Later, mice were tested to assess the memory strength for the context-shock relationship by placing them in the same chamber, but without sounding the cue. In this situation, wild-type mice freeze because they recognize the surroundings and associate them with the electric shock. Both male and female ßERKOWYE mice were significantly impaired in their freezing response to the context—indicative of failed associative learning of the shock with the context. It is thought that that memory encoding and retrieval of a context is dependent on the hippocampus. It is well established that genetic manipulations that disrupt hippocampal memory also disrupt forms of hippocampal synaptic plasticity, such as long-term potentiation. Interestingly, hippocampal slices taken from female ßERKOWYE exhibit a robust long-term potentiation deficit in area CA1 compared with wild-type mice. Specifically, fast synaptic transmission was impaired and long-term potentiation responses were blunted, and both of these deficits would be expected to impact learning and memory.
Models of Inflammatory Bowel Disease and Arthritis
Nonselective estrogens are known to have antiinflammatory properties and to interfere with NF
B activity (72, 73) and thus it was possible that ERß might be capable of mediating this activity. Many models of inflammation exist, but the first model we used was the HLA-B27 transgenic rat. These rats develop inflammation in the intestine, joint, and skin and are used as models of inflammatory bowel disease and rheumatoid arthritis (74). When administered to fully diseased rats, ERB-041 and WAY-202196 reversed the animals’ symptoms of chronic diarrhea and reduced intestinal inflammation (25, 41). Coadministration of ICI-182780 completely blocked the effect of ERB-041 on stool character, providing strong evidence that its effect was ER dependent. At least seven other ERß-selective agonists, representing several different chemical series, also have this effect (Ref. 24 ; and Harris, H. A., and J. C. Keith, Jr., unpublished observations). In addition, ERB-041 and WAY-202196 prevent or reverse established arthropathy in these rats as reflected by a reduction in the clinical signs of joint redness/swelling as well as synovitis and Mankin histological scores (Refs. 24 and 25 ; and Harris, H. A., and J. C. Keith, Jr., unpublished observations).
ERB-041 and WAY-202196 were also tested in the Lewis rat adjuvant-induced arthritis model, a well-established model of rheumatoid arthritis. Both these compounds reduced clinical signs of arthritis to near normal levels and significantly improved joint histology scores. In addition, ERB-041 reversed the majority of gene expression changes caused by disease induction in the liver, lymph node, and spleen and, as well, reversed many of the disease-induced plasma proteome changes (75).
Although nonselective estrogens/selective ER modulators are active in both these models (76, 77), ER-selective compounds have not been tested. Moreover, these observations cannot be confirmed using ERKO mice because they rely on specific genetic models of disease. Our experience with Wyeth ERß-selective agonists is that they are not general antiinflammatory agents because they are inactive in many other models of inflammation, including collagen-induced arthritis, dextran sulfate sodium-induced colitis, carrageen-induced paw edema, lipopolysaccharide-induced haptoglobin production, experimental allergic encephalitis, and passive cutaneous anaphalaxis (Ref. 78 and Harris, H.A., unpublished observations). Finally it should be mentioned that a clear role for ER-dependent antiinflammatory activity has been established in some in vivo models; thus, antiinflammatory activity is not a unique property to ERß (78, 79, 80, 81).
Endometriosis
Endometriosis is an estrogen-dependent disease with a significant inflammatory component (82). Although it may seem counterintuitive to evaluate an ER agonist in models of this disease, given the antiinflammatory activity seen in other models, it seemed a reasonable idea to pursue. Although this disease occurs naturally only in primates, several rodent models exist, and one well-accepted model uses human endometrial tissue implanted as xenografts into nude mice (83, 84). ERB-041 was tested in both ovariectomized and gonad intact mice with established endometriosis lesions, and after 2 wk of treatment, 40–75% of the mice were completely lesion free (85). Only ER was detected in recovered lesions (regardless of treatment), and thus it is speculated that ERB-041 may be stimulating the macrophages and/or natural killer cells of the nude mice to recognize the exogenous endometrial tissue as foreign and clear it. Clearly the in vivo profile of a selective ERß agonist is not only a subset of estradiol’s activity.
Trauma and Sepsis Models
It has become apparent in recent years that females have a lower susceptibility for, and an improved survival from, traumatic injury and sepsis (86). There is a large body of work showing androgen’s deleterious effects and estrogen’s protective qualities. Recent work is beginning to delineate the contributions of ER and ERß to this beneficial effect. For the sake of brevity, only data relating to ERß will be presented here. In a rat model of trauma-hemorrhage, which examines lung injury, estradiol and DPN were equally effective in reducing a variety of inflammatory mediators including inducible nitric oxide synthase and IL-6, as well as markers of tissue damage such as lactate dehydrogenase and total protein content of the bronchioalveolar fluid. PPT was ineffective on these parameters (87). In a similar study, designed to examine cardiac function, DPN, but not PPT, improved cardiac output and stroke volume. In addition DPN, but not PPT, maintained heat shock proteins 32, 60, 70, and 90 (88).
WAY-202196 has been examined in three models of systemic infection, the second and third of which are used as models of sepsis: 1) mouse listeriosis, 2) Pseudomonas infection in the neutropenic rat, and 3) mouse cecal ligation and puncture (89). Whereas a single oral dose of WAY-202196 did not positively or negatively affect response to Listeria infection, a monoclonal antibody to TNF increased susceptibility to infection (lowered the LD50). Thus, WAY-202196 was not globally immunosuppressive in this model.
The Pseudomonas infection model aims to replicate the human situation of an immunocompromised patient (e.g. cancer patient receiving radio- or chemotherapy) who develops an infection. In this model, multiple doses of WAY-202196 increased survival (25–83%) and improved intestinal histology. The mouse cecal ligation and puncture model induces acute bacterial peritonitis (for review, see Ref. 90). When WAY-202196 was tested in this model, it improved survival (0% vs. 83%) and also reduced intestinal injury.
Leukemia
At 18 months of age, ßERKOKI mice of both sexes have grossly enlarged spleens compared with their wild-type counterparts, and is thought to be a consequence of bone marrow hypercellularity (91). Moreover, these mice develop neoplasia resembling human chronic myeloid leukemia. To date, this phenotype has not been reported in other strains of ßERKO mice.
|
UNANSWERED QUESTIONS ABOUT ERß |
|---|
|
TOPABSTRACTINTRODUCTIONSUMMARY OF AVAILABLE TOOLS...ER{alpha} MEDIATES THE MAJORITY...ROLES FOR ERß IN...UNANSWERED QUESTIONS ABOUT...CONCLUDING REMARKSREFERENCES |
|---|
to ameliorate hot flushes and bone loss (despite having known about these activities for decades), mechanistic uncertainty undoubtedly handicaps further research. It seems doubtful that it will be possible to explain the diverse effects of this transcription factor with a single mechanism. Having said that, however, one system that bears further investigation is the hypothalamic-pituitary-adrenal (HPA) axis. Modulation of this axis may explain effects in certain inflammatory and behavioral models. First, looking at expression in relevant brain regions, ERß is the predominant ER expressed in the rat paraventricular nucleus (10), which is a key mediator of the stress response. Additionally, there is evidence for an interplay between corticosterone, estradiol, and ERß expression in the paraventricular nucleus (92). Finally, ERß is colocalized with a subset of CRH neurons in the rat paraventricular nucleus and in vitro, ERß can up-regulate the CRH promoter in response to estradiol (93).
Several studies demonstrate an effect of nonselective estrogens on the HPA. It should be noted that effectiveness on these endpoints is highly dependent on dose and regimen. For the sake of brevity, only a few studies will be mentioned here. A single episode of restraint stress markedly elevated ACTH and corticosterone levels in rats, but chronic treatment with estradiol benzoate significantly reduced the ACTH response (with no effect on corticosterone levels) (94). Interestingly, in this study, estradiol benzoate increased levels of other stress-responsive markers (adrenal tyrosine hydroxylase and dopamine ß-hydroxylase mRNA) in unstressed animals, pointing out the system’s complexity. In another stress-dependent model, estradiol reduced stress-induced expression of c-Fos in the paraventricular nucleus, although (as above) corticosterone levels were unaffected (95).
Other studies have measured behavioral consequences of HPA modulation. In a study discussed above, Lund et al. (70) showed that all rats tested in the elevated plus maze had increased levels of corticosterone, but the levels seen in the DPN-treated animals were significantly lower than those treated with estradiol or PPT (which were significantly higher than vehicle-treated rats in contrast to the findings of Ref. 96). Because both estradiol and DPN were effective at reducing anxiety in this model, the effects would seem not explainable by modulation of corticosterone levels, although it should be noted that it was only measured at one time point (30 min after returning to home cage).
In a similar study again using the open field and elevated plus maze tests, Walf and Frye (96) found that estradiol’s effects in reducing anxiety were blunted by pretest restraint stress. Doses of estradiol that were effective in reducing stress also reduced corticosterone levels and these effects were also seen in unstressed untested rats. Estradiol’s antianxiety action was likely not entirely via modulation of corticosterone levels, because although estradiol was ineffective in adrenalectomized rats, this was reversed by administration of low doses (but not high doses) of corticosterone.
Lewis rats, which are used for the adjuvant-induced arthritis model, have well-known defects in their HPA axis that render them hypersensitive to inflammatory stimuli (97). Lipopolysaccharide stimulation of Fischer 344 rats (the control strain for Lewis rats) leads to robust up-regulation of CRH mRNA in the paraventricular nucleus and a modest down-regulation of ER and ERß mRNA (97). Neither of these responses is seen in Lewis rats, again suggesting interplay between ERs and the HPA.
Are There Expected Clinical Liabilities with Selective ERß Agonists?
Based on the biological profile of selective ERß agonists to date, are there any obvious clinical liabilities that might hinder their development? Answering this question, again, requires rigorous testing of a variety of compounds to separate observations based on mechanism from those that might be idiosyncratic. Given the widespread expression of ERß, it was surprising to us that we could find essentially no ERß-dependent genes in normal (healthy) mice dosed with estradiol (98). This observation contributed to the construction of the challenge hypothesis, i.e. that ERß’s function becomes apparent only when the animal is stressed, ill, or otherwise compromised. If this is true, then effects of ERß activation in normal tissues may not lead to many responses and will afford less opportunity for unwanted side effects.
One very difficult but critical question to answer is the influence of ERß agonists on the risk of developing breast cancer in humans. Preclinically, proliferative effects of estrogens on human breast cancer cells in vitro or in animal models of breast cancer is easily demonstrated. For example, a number of human breast cancer cell lines proliferate in response to estrogens, and antiestrogens or selective ER modulators, which are clinically useful agents to treat human cancers, block this effect. In addition, transgenic mice expressing the ER coactivator amplified in breast cancer 1/steroid receptor coactivator 3 develop mammary tumors whereas null mice are resistant (99, 100). These studies have primarily focused on the role of ER in mediating the effects of estrogens.
It should be remembered however, that, even though many estrogen-responsive preclinical models of breast cancer exist, a large-scale clinical trial involving more than 10,000 women taking 0.625 mg/d of conjugated equine estrogens for a mean of 6.8 yr demonstrated that this regimen did not increase the risk for developing breast cancer (101). In fact, there was a trend toward a decreased risk (hazard ratio with 95% confidence interval = 0.59–1.01). This observation demonstrates that the effects of estrogens seen in preclinical models of breast cancer do not wholly translate to humans.
ERß is expressed in the human mammary gland, and many studies have tried to define its role. The goal of mapping studies is to correlate expression of ER and ERß with clinical parameters. It should be cautioned, however, that a correlation between two observations does not guarantee that they are causally related. Many such studies have been conducted, and the results from those published from 1997–2003 (mRNA) or 1999–2003 (protein) are summarized in a recent review (102), and the conclusions are concordant with work published more recently (103, 104, 105, 106, 107, 108, 109). Regarding mRNA expression, the majority of these studies indicate that healthy mammary glands express more ERß mRNA than do breast cancer samples. ERß expression has been correlated with tamoxifen sensitivity and disease-free/overall survival. In breast cancer samples, ER is usually the predominant ER, and ERß expression has been reported to decrease during carcinogenesis. Promoter methylation has been implicated as a cause of reduced ERß expression (110), and a decrease in ERß expression has also been reported for other cancers such as prostate, endometrium, ovary, and colon and has led to the hypothesis that ERß is a potential tumor suppressor (see Ref. 111 for review).
A number of in vitro functional studies have been performed to examine the effect of ERß expression on the proliferation of breast cancer cells (112, 113, 114, 115, 116). Although the results are not unanimous, the majority of studies conclude that an increase in ERß expression decreases cell proliferation. Additionally, when ligand dependence of the effects was evaluated, most studies agree that the effects seen are ligand independent. Therefore, they may have minimal implication for predicting the activity of an ERß-selective agonist. Taken together, the current data do not support a role for ERß’s ligand-dependent activity in the genesis or progression of breast cancer.
|
CONCLUDING REMARKS |
|---|
|
TOPABSTRACTINTRODUCTIONSUMMARY OF AVAILABLE TOOLS...ER{alpha} MEDIATES THE MAJORITY...ROLES FOR ERß IN...UNANSWERED QUESTIONS ABOUT...CONCLUDING REMARKSREFERENCES |
|---|
Note Added in Proof
Since this manuscript was accepted, data using PPT, ERB-041 and WAY-202196 have been published showing that ER stimulation, in combination with progesterone, is necessary and sufficient to elicit changes on mouse mammary gland morphology and regulation of at least one mRNA marker of hormone response (117).
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
First Published Online March 23, 2006
Abbreviations: DPN, Diarylpropionitrile; ER, estrogen receptor; ERKO, ER knockout; ßERKOCH, mice produced in Chapel Hill; ßERKOKI, mice established at Karolinska Institute; ßERKOST, mice produced at Strasbourg, France; ßERKOWYE, mice produced at Wyeth; ERB-041, 2-(3-fluoro-4-hydroxyphenyl)-7-vinyl-1,3-benzoxazol-5-ol; HPA, hypothalamic-pituitary-adrenal; KO, knockout; 16-LE2, 3,17-dihydroxy-19-nor-17-pregna-1,3,5(10)-triene-21,16-lactone; PPT, propylpyrazole triol; 8ß-VE2, 8-vinylestra-1,3,5(10)-triene-3,17ß-diol; WAY-200070, 7-bromo-2-(4-hydroxyphenyl)-1,3-benzoxazol-5-ol; WAY-202196, 3-(3-fluoro-4-hydroxy-phenyl)-7-hydroxy-naphthonitrile.
Received for publication November 17, 2005. Accepted for publication March 16, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONSUMMARY OF AVAILABLE TOOLS...ER{alpha} MEDIATES THE MAJORITY...ROLES FOR ERß IN...UNANSWERED QUESTIONS ABOUT...CONCLUDING REMARKSREFERENCES |
|---|
and ß. Endocrinology 138:863–870 and ß are species dependent. Steroids 67:379–384[CrossRef][Medline] and -ß mRNA in the rat central nervous system. J Comp Neurol 388:507–525[CrossRef][Medline] (ER-) and ß (ER-ß) mRNA in the rat pituitary, gonad, and reproductive tract. Steroids 63:498–504[CrossRef][Medline] (ER) and estrogen receptor-ß (ERß) messenger ribonucleic acid in the wild-type and ER-knockout mouse. Endocrinology 138:4613–4621 (ER ) and ß (ER ß) on mouse reproductive phenotypes. Development 127:4277–4291[Abstract] and ß with potent selective ligands from structure-based design. Mol Endocrinol 18:1599–1609-selective agonists. J Med Chem 43:4934–4947[CrossRef][Medline] and ER ß, in estrogen target tissues in vivo through the use of an ER -selective ligand. Endocrinology 143:4172–4177 agonist 16 -LE2 inhibits cardiac hypertrophy and improves hemodynamic function in estrogen-deficient spontaneously hypertensive rats. Cardiovasc Res 67:604–612 in the uterus. Proc Natl Acad Sci USA 97:5936–5941 is required for mammary development and the induction of mammary hyperplasia and epigenetic alterations in the aromatase transgenic mice. J Steroid Biochem Mol Biol 95:9–15[CrossRef][Medline] and ER ß on progesterone receptor expression and action and on gonadotropin and prolactin secretion in the rat. Neuroendocrinology 79:247–258[CrossRef][Medline] and ß to protect hippocampal neurons against global ischemia-induced cell death. Endocrinology 146:3070–3079 activity by ER ß in the uterus. Endocrinology 144:3159–3166: studies with aERKO and BERKO mice. Cancer Res 61:6089–6097 and imprinting of the neonatal mouse ventral prostate by estrogen. Proc Natl Acad Sci USA 102:1484–1489 and ß in reproductive tissues. J Steroid Biochem Mol Biol 74:287–296[CrossRef][Medline]-androstane-3 ß,17 ß-diol, and CYP7B1, regulates prostate growth. Proc Natl Acad Sci USA 99:13589–13594 or ERß. Endocrinology 147:191–200KO and ERßKO mice. Matuitas 53:210–216 mediates the protective effects of estrogen against vascular injury. Circ Res 90:1087–1092 agonists and is abolished by estrogen deprivation. J Pharmacol Exp Ther 313:1203–1208 stimulation in ovariectomized SHR. Hypertension 42:991–996, not ß, is a critical link in estradiol-mediated protection against brain injury. Proc Natl Acad Sci USA 98:1952–1957 and ER-ß in the midbrain raphe nuclei and periaqueductal gray in male mouse: predominant role of ER-ß in midbrain serotonergic systems. Neuroscience 130:445–456[CrossRef][Medline] B in human coronary smooth muscle cells. Circ Res 87:1006–1011 B activity in vivo. Circ Res 89:823–830. Am J Pathol 163:1599–1605 mediates the brain anti inflammatory activity of estradiol. Proc Natl Acad Sci USA 100:9614–9619 signaling in inflammatory leukocytes is dispensable for 17 ß-estradiol-mediated inhibition of experimental autoimmune encephalomyelitis. J Immunol 173:2435–2442 regulate corticotropin-releasing hormone promoter activity through an alternate pathway. J Neurosci 24:10628–10635 and ß profiling in human breast cancer. Eur J Surg Oncol 30:469–474[CrossRef][Medline] function differently in breast cancer cell line MCF7. Oncogene 22:5011–5020[CrossRef][Medline]NURSA Molecule Pages Link:
This article has been cited by other articles:
 |
|
 |
|
  C. Carli, C. N. Metz, Y. Al-Abed, P. H. Naccache, and A. Akoum Up-Regulation of Cyclooxygenase-2 Expression and Prostaglandin E2 Production in Human Endometriotic Cells by Macrophage Migration Inhibitory Factor: Involvement of Novel Kinase Signaling Pathways Endocrinology, July 1, 2009; 150(7): 3128 - 3137. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. J. Weiser, T. J. Wu, and R. J. Handa Estrogen Receptor-{beta} Agonist Diarylpropionitrile: Biological Activities of R- and S-Enantiomers on Behavior and Hormonal Response to Stress Endocrinology, April 1, 2009; 150(4): 1817 - 1825. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Trukhacheva, Z. Lin, S. Reierstad, Y.-H. Cheng, M. Milad, and S. E. Bulun Estrogen Receptor (ER) {beta} Regulates ER{alpha} Expression in Stromal Cells Derived from Ovarian Endometriosis J. Clin. Endocrinol. Metab., February 1, 2009; 94(2): 615 - 622. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Cignarella, C. Bolego, V. Pelosi, C. Meda, A. Krust, C. Pinna, R. M. Gaion, E. Vegeto, and A. Maggi Distinct Roles of Estrogen Receptor-{alpha} and {beta} in the Modulation of Vascular Inducible Nitric-Oxide Synthase in Diabetes J. Pharmacol. Exp. Ther., January 1, 2009; 328(1): 174 - 182. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. R. Bhavnani, S.-P. Tam, and X. Lu Structure Activity Relationships and Differential Interactions and Functional Activity of Various Equine Estrogens Mediated via Estrogen Receptors (ERs) ER{alpha} and ER{beta} Endocrinology, October 1, 2008; 149(10): 4857 - 4870. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Baylis Sexual Dimorphism of the Aging Kidney: Role of Nitric Oxide Deficiency Physiology, June 1, 2008; 23(3): 142 - 150. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. Guzeloglu-Kayisli, G. Halis, S. Taskiran, U. A. Kayisli, and A. Arici DNA-Binding Ability of NF-{kappa}B is Affected Differently by ER{alpha} and ER{beta} and Its Activation Results in Inhibition of Estrogen Responsiveness Reproductive Sciences, May 1, 2008; 15(5): 493 - 505. [Abstract] [PDF]
|
|
|
|
 |
|
 E. C. Chang, T. H. Charn, S.-H. Park, W. G. Helferich, B. Komm, J. A. Katzenellenbogen, and B. S. Katzenellenbogen Estrogen Receptors {alpha} and {beta} as Determinants of Gene Expression: Influence of Ligand, Dose, and Chromatin Binding Mol. Endocrinol., May 1, 2008; 22(5): 1032 - 1043. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. L. Yanes, J. C. Sartori-Valinotti, and J. F. Reckelhoff Sex Steroids and Renal Disease: Lessons From Animal Studies Hypertension, April 1, 2008; 51(4): 976 - 981. [Full Text] [PDF]
|
|
|
|
|
|
O. Bukulmez, D. B. Hardy, B. R. Carr, R. A. Word, and C. R. Mendelson Inflammatory Status Influences Aromatase and Steroid Receptor Expression in Endometriosis Endocrinology, March 1, 2008; 149(3): 1190 - 1204. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. C. Antal, A. Krust, P. Chambon, and M. Mark Sterility and absence of histopathological defects in nonreproductive organs of a mouse ER{beta}-null mutant PNAS, February 19, 2008; 105(7): 2433 - 2438. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Cvoro, D. Tatomer, M.-K. Tee, T. Zogovic, H. A. Harris, and D. C. Leitman Selective Estrogen Receptor- Agonists Repress Transcription of Proinflammatory Genes J. Immunol., January 1, 2008; 180(1): 630 - 636. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. D. Gardner, G. L. Brower, T. G. Voloshenyuk, and J. S. Janicki Cardioprotection in female rats subjected to chronic volume overload: synergistic interaction of estrogen and phytoestrogens Am J Physiol Heart Circ Physiol, January 1, 2008; 294(1): H198 - H204. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Q. Xue, Z. Lin, Y.-H. Cheng, C.-C. Huang, E. Marsh, P. Yin, M. P Milad, E. Confino, S. Reierstad, J. Innes, et al. Promoter Methylation Regulates Estrogen Receptor 2 in Human Endometrium and Endometriosis Biol Reprod, October 1, 2007; 77(4): 681 - 687. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Silvestri and H. M Fraser Oestrogen and progesterone receptors in the marmoset endometrium: changes during the ovulatory cycle, early pregnancy and after inhibition of vascular endothelial growth factor, GnRH or ovariectomy Reproduction, August 1, 2007; 134(2): 341 - 353. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Heldring, A. Pike, S. Andersson, J. Matthews, G. Cheng, J. Hartman, M. Tujague, A. Strom, E. Treuter, M. Warner, et al. Estrogen Receptors: How Do They Signal and What Are Their Targets Physiol Rev, July 1, 2007; 87(3): 905 - 931. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
